From October through December 2021, a cross-sectional survey was conducted in Chengdu, a provincial capital in Southwest China. It was mainly designed to assess the prevalence of and risk factors for psychobehavioral problems among preschool children. A total of 4360 preschool children from 30 kindergartens, were selected by cluster random sampling method. The study protocol was approved by the Ethics Committee of Soochow University (Approval NO. SUDA20210820H01). Written informed consent was obtained from each of the participants (guardians).

In the present study, we excluded 129 children aged less than 3 years and 34 children with missing information. Therefore, the final study population was 4197 children (2190 boys and 2007 girls, mean age 4.28 ± 0.93 years).

Data were collected by a validated questionnaire administered by guardians via WJX (an online survey platform, www.wjx.cn). Information on two independent variables was obtained by the following questions: “How often did your child sleep with the lamplight on at night? (a) 0 night/week, (b) 1-2 nights/week, (c) ≥3 nights/week”, and “For how long did the child sleep daily including siesta? (a) <10 hours/day, (b) 10-12 hours/day, (c) >12 hours/day”. According to the National Sleep Foundation’s sleep duration recommendation (16), total sleep duration with <10 hours/day for preschool children was defined as insufficient sleep in our study.

ADHD was screened using the Conners Parental Symptom Questionnaire (PSQ), which was compiled originally by the American scholar Conners and revised by Conners, Goyette, and Ulrich in 1978 (17). The PSQ has been widely applied to evaluate psychobehavioral problems of children and adolescents aged 3-17 years. It has been introduced or amended by many countries and shown high reliability and validity (18–20). Cronbach’s α value was 0.933 in our survey.

The 1978 version of PSQ was used in the present study, which included a total of 48 items and 6 measurement dimensions (i.e., conduct problems, learning problems, psychosomatic disorders, hyperactivity-impulsivity, anxiety, hyperactivity index). Hyperactivity index including 10 items was a special screening tool for ADHD. Participants were asked to rate their child’s behavior performance in the past six months. Each item was scored from 0 (not at all) to 3 (very often). ADHD was defined as having an average score of hyperactivity index of ≥1.5.

The children’s demographic characteristics including age (continuous), sex (male or female), lifestyle and family factors including snacks intake (0 days/week, 1-2 days/week, 3-4 days/week or ≥5 days/week), picky eating (yes or no), time of watching TV (0 hours/day, <1 hours/day, 1-2 hours/day or >2 hours/day), home lighting (dark, ordinary, or bright) were collected via the questionnaire. In addition, perinatal factors of the child’s mother including smoking or passive smoking during pregnancy (0 days/week, 1-2 days/week or ≥3 days/week), exercise during pregnancy (0 minutes/day, <20 minutes/day, 20-40 minutes/day, or >40 minutes/day), gestational hyperglycemia (yes or no), gestational hypertension (yes or no), gestational anemia (yes or no), and anxiety or depression during pregnancy (yes or no) were also collected.

The basic characteristics of the participants by ADHD status, lamplight exposure during sleep at night and different sleep durations are presented as numbers (percentages), and were compared using Chi-square test or Fisher’s exact test.

We used logistic regression to evaluate the associations of lamplight exposure during sleep at night and sleep duration with ADHD. Three multivariable models were constructed: model 1 was the basic model adjusted for age and sex; model 2 was adjusted as model 1 plus the child’s lifestyle and family factors; model 3 was adjusted as model 2 plus the perinatal factors of mothers. P-trend was calculated by modeling the independent variables as continuous ones in the regression models.

Effect modification of the association between lamplight exposure during sleep and ADHD by sex, picky eating, gestational hyperglycemia and gestational anemia was estimated in stratified analyses, and significance of interactions was evaluated on first-degree multiplicative models for each stratification variable separately.

The statistical analyses were performed using SPSS 26.0 statistical software (IBM Corp., USA). All tests were two-tailed, and a P value <0.05 was considered statistically significant.

In this cross-sectional study, we observed that lamplight exposure during sleep at night was associated with increased odds of ADHD among the preschool children. Moreover, both sleep duration of <10 hours/day and >12 hours/day increased the likelihood of ADHD as well.

Accumulating evidence supports that exposure to lamplight (indoor and outdoor light) in the evening is related to various adverse health problems including psychobehavioral disorders (21–24). Outdoor light exposure at night was significantly associated with a higher risk of autism spectrum disorder (ASD, a neurodevelopmental disorder) in children (24); nighttime exposure to light could be able to induce neurological changes, mood disorders and depressive-like behaviors (21–24).

In our study, insufficient sleep duration increased odds of ADHD, which was consistent with previous studies (11, 12). Long sleep duration also increased the likelihood of ADHD, which was supported by Bogdan’s findings (13), but inconsistent with Lee’s (17). The inconsistency may come from different study design, screening tools for ADHD, classification of sleep duration, sample size and characteristics of the study populations.

The underlying mechanisms for the association of lamplight exposure during sleep with ADHD remain unclear. However, we proposed several potential explanations.

First, nighttime light can directly affect cognition and mood through intrinsically photosensitive retinal ganglion cells (ipRGC) projections to brain regions involved in emotional regulation (25). Second, nighttime light can indirectly affect mood and behaviors by altering clock and inflammation genes expression, hormone secretion, neuroplasticity and neurotransmission (22). Third, nighttime lamplight is able to disturb the circadian rhythm, inhibit the secretion of melatonin, directly stimulate brain activity (26, 27), and consequently trigger sleep disorders, which have been associated with the risk of ADHD (28, 29). Additionally, studies have demonstrated that light exposure can alter gene expression linked to (neuro)development, including crf, crfbp, mr, and gr-alpha (30). Crf antagonists improved the ability to work memory, the same as all approved ADHD treatments (31). This suggests that nighttime lamplight exposure in preschool children might raise ADHD risk potentially via crf gene regulation.

In this study, we also observed positive associations of both short and long sleep duration with ADHD among the preschool children. Insufficient sleep could damage the prefrontal cortex of the brain and limit the link between the prefrontal cortex and the amygdala (32), thereby affecting cognition, emotion control and executive function (33), which may explain the neurobehavioral, neurocognitive, and functional manifestations of ADHD symptomatology (28). The underlying mechanisms for the association of excessive sleep with ADHD remain unclear, the following mechanism may be involved: excessive sleep may increase arterial stiffness (34), reduce blood flow, and thus adversely affect the nervous system of the children.

To the best of our knowledge, this is the first study to examine the association between lamplight exposure during sleep and ADHD. Other main strengths include a relatively large sample size and adjustment for a number of confounders in the regression models.

Nevertheless, the study has also several limitations. First, this study was based on a cross-sectional design, therefore causal and temporal relationships could not be inferred. Reverse causality (i.e. ADHD causes sleeping with the light on at night and changed sleeping duration) can not be precluded. Researchers should confirm causality through longitudinal study and experimental study in the future. Second, based on the PSQ scale, preschool children with a hyperactivity index greater than 1.5 were classified as ADHD in this study. This may have resulted in misclassification. To address this limitation, future studies should consider using clinically diagnosed ADHD cases to enhance accuracy. Third, reporting bias and recall bias might have occurred due to self-report by guardians of the children. Future studies should prefer using objective measures (e.g., actigraphy for sleep). Fourth, although we have adjusted for a number of confounders, residual confounding by unknown or unmeasured factors including other sleep disorders cannot be ruled out. For example, children with low socioeconomic status may have a greater likelihood of ADHD (35). A central disorder of hypersomnolence may cause prolonged nighttime sleepiness, and studies have shown that ADHD occurs with a high frequency in adults with this condition (36). Fifth, our study focused on preschool children in Chengdu city, which may have restricted generalizability of the results. Therefore, future studies from different geographic areas or age groups are warranted.